Project Summary: This proposal uses a multidimensional approach for target validation and drug discovery in hepatocellular carcinoma (HCC) that focuses on pharmacogenomic predictions, a proprietary chemical analog library built on approved HCC KI scaffolds, and the use of multiple precision models of HCC: 1) Genetically defined murine HCCs induced in a fully immunocompetent background. This will allow for in vivo drug testing and to assess the drug impact in the context of the tumor microenvironment and in combination with checkpoint inhibitors; 2) Three-dimensional tumor organoids derived from these same murine models, which allow scalability for drug screening. Tumor organoids are a major breakthrough for convenient omics-based analyses of tumor biology and preclinical drug discovery, and are shown to accurately recapitulate patient responses to anticancer agents; 3) Patient derived organoids and 4) Patient derived xenografts (PDX), which allow testing of drug efficacy in genetically complex primary human tumors. Our close collaborator, Arvin Dar, has successfully applied aspects of this approach to kinases involved in medullary thyroid carcinoma (Dar et al., Nature, 2012; Sonoshita et al., Nature Chemical Biology, 2018). We have also recently applied a simplified approach and confirmed increased anti-tumoral activity of a new KI -AD80- compared to the standard-of-care (sorafenib) in experimental HCC models (Yu et al, accepted). To date, no single study has combined the multidisciplinary innovation presented in this proposal for drug discovery in liver cancer. The major hypothesis of this project is that different HCC oncogenic drivers will establish unique therapeutic vulnerabilities within tumor organoid lines. By using multiple precision models and informed modifications to KIs, we will be able to suggest stratification strategies and identify better therapeutics for HCC patients. This rationale is currently unprecedented in liver cancer research.